Two kinds of nested structures that are supported: arrays and maps. Fields of either of these types are arrays; when they are part of a query result, which is already an array of objects, they become nested arrays.
The GraphQL API supports two different sorts of arrays:
- Some arrays contain scalar types (for example,
[String]) and function like ordinary fields that can be filtered - Some arrays contain more complex types (for example,
[Subrequest].) The following section describes their behaviour.
Arrays of non-scalar types behave as a single value. There is no way to paginate through, filter, filter by, group, or group by the array.
On the other hand, you can choose which fields of the underlying type you want fetched.
For example, given arrays like this:
type SubRequest {
url: String!
status: Int
}
type Request {
date: Date!
datetime: DateTime!
subRequests: [SubRequest!]!
}You can run a query to get the status by subrequest:
{
requests {
date
subRequests {
# discard the url, only need the status
status
}
}
}The results would be:
{
"requests": [
{
"date": "2018-01-01",
"subRequests": [{"status": 404}, {"status": 200}, {"status": 404}]
},
{
"date": "2018-01-01",
"subRequests": [{"status": 200}]
}
]
}Maps behave like arrays, but can be grouped using the sum function. They are used in aggregated datasets, such as httpRequest1dGroups.
Example maps:
type URLStatsMapElem {
url: String!
requests: Int
bytes: Int
}
type Request {
date: Date!
datetime: DateTime!
urlStatsMap: [URLStatsMapElem!]!
}Query:
{
requests {
sum {
urlStatsMap {
url
requests
bytes
}
}
dimensions {
date
}
}
}Response:
{
"requests": [
{
"sum": {
"urlStatsMap": [
{
"url": "hello-world.org/1",
"requests": 123,
"bytes": 1024
},
{
"url": "hello-world.org/10",
"requests": 1230,
"bytes": 10240
}
]
}
"dimensions" {
"date": "2018-10-19"
}
},
...
]
}Query array fields in raw datasets:
query NestedFields($zoneTag: string, $start: Time, $end: Time) {
viewer {
zones(filter: { zoneTag: $zoneTag }) {
events(limit: 2, filter: { datetime_geq: $start, datetime_leq: $end }) {
matches {
ruleId
action
source
}
}
}
}
}Example response:
{
"data": {
"viewer": {
"zones": [
{
"events": [
{
"matches": [
{
"action": "allow",
"ruleId": "rule-id-one",
"source": "asn"
},
{
"action": "block",
"ruleId": "rule-id-two",
"source": "asn"
}
]
}
]
}
]
}
},
"errors": null
}Query maps fields in aggregated datasets:
query MapCapacity(
$zoneTag: string
$dateStart: Date
$dateEnd: Date
$start: Time
$end: Time
) {
viewer {
zones(filter: { zoneTag: $zoneTag }) {
httpRequests1mGroups(
limit: 10
filter: {
date_geq: $dateStart
date_leq: $dateEnd
datetime_geq: $start
datetime_lt: $end
}
) {
sum {
countryMap {
clientCountryName
requests
bytes
threats
}
}
dimensions {
datetimeHour
}
}
}
}
}Example response:
{
"data": {
"viewer": {
"zones": [
{
"httpRequests1mGroups": [
{
"dimensions": {
"datetime": "2019-03-08T17:00:00Z"
},
"sum": {
"countryMap": [
{
"bytes": 51911317,
"clientCountryName": "XK",
"requests": 4492,
"threats": 0
},
{
"bytes": 1816103586,
"clientCountryName": "T1",
"requests": 132423,
"threats": 0
},
...
]
}
}
]
}
]
}
},
"errors": null
}